Abstract
Abstract This paper presents a set of control approaches that can deal with unwanted dynamics such as the Hopf Bifurcation in a Chaotic Associative Memory (CAM) for two distinct goals: the bifurcation critical point shift to anticipate or to postpone the start of a Hopf bifurcation; or to extinguish it. A Washout Filter Control, a Polynomial Control, and an adaptive control for data-driven called the Model-Free Adaptive Control were used in this study. We present the results of numerical simulations in a CAM with three neurons (a system with six dimensions), which were trained classically to retrieve a set of memories. Our analysis includes the dynamics of the bifurcated system, the ability of the network to recover the memories learned and how it is affected by the control strategies. The generality of the proposed control approaches was tested to control a different system and for a different behavior (chaotic dynamics).
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call